Object Detection on MSCOCO 2014 (test)
29.1mAP@.5Object Discovery via Contrastive Learning
Evaluation results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Object Discovery via Contrastive LearningBackbone=ResNet502022.08 | 29.1 | 13.9 | 11.8 | 4.9 | 16.8 | 22.3 | 15.5 | 26.1 | 28 | 9 | 31.8 | 46.6 | |
| Object Discovery via Contrastive LearningBackbone=ResNet1012022.08 | 29 | 14.4 | 12.4 | 4.8 | 17.3 | 23.8 | 15.8 | 27 | 30 | 9.2 | 33.6 | 51 | |
| CASDBackbone=ResNet502022.08 | 27.8 | 13 | — | — | — | — | — | — | — | — | — | — | |
| Object Discovery via Contrastive LearningBackbone=VGG162022.08 | 27.7 | 13.7 | 11.9 | 4.4 | 14.5 | 21.2 | 14.7 | 24.8 | 26.9 | 8.8 | 27.8 | 44 | |
| CASDBackbone=VGG162022.08 | 26.4 | 12.8 | — | — | — | — | — | — | — | — | — | — | |
| MISTBackbone=ResNet502022.08 | 26.1 | 12.6 | 10.8 | 3.7 | 13.3 | 19.9 | 14.8 | 23.7 | 24.7 | 8.4 | 25.1 | 41.8 | |
| MISTBackbone=ResNet1012022.08 | 26.1 | 13 | 10.8 | 3.7 | 13.3 | 19.9 | 14.8 | 23.7 | 24.7 | 8.4 | 25.1 | 41.8 | |
| MISTBackbone=VGG162022.08 | 24.3 | 11.4 | 9.4 | 3.6 | 12.2 | 17.6 | 13.5 | 22.6 | 23.9 | 8.5 | 25.4 | 38.3 | |
| WSOD2Backbone=VGG162022.08 | 22.7 | 10.8 | — | — | — | — | — | — | — | — | — | — | |
| C-MIDNBackbone=VGG162022.08 | 21.4 | 9.6 | — | — | — | — | — | — | — | — | — | — | |
| PCLBackbone=VGG162022.08 | 19.4 | 8.5 | — | — | — | — | — | — | — | — | — | — | |
| MELMCNN=VGG162019.02 | 18.8 | 7.8 | — | — | — | — | — | — | — | — | — | — | |
| MELMCNN=VGGF2019.02 | 11.9 | 4.1 | — | — | — | — | — | — | — | — | — | — | |
| WSDDNCNN=VGGF2019.02 | 10.1 | 3.1 | — | — | — | — | — | — | — | — | — | — |